In the year following the implementation of our bar code medication administration (BCMA) system, we experienced an increase in reported medication errors, which, while common with new system implementations, was nonetheless a cause for grave concern. We noticed that staff’s tendency to blame the technology when problems arose prevented individuals from taking responsibility for finding solutions. Our goal was to inspire our staff to examine the sources of problems and devise solutions to overcome not only technology malfunctions, but also barriers in system design and use.1 Undertaking an exhaustive failure modes effects analysis (FMEA) provided the tools to create a successful error-reduction process.
Cause for Concern: Increased Errors
We first implemented BCMA at our three-facility health system over a few months in 2008, and in August, 2009, we noticed an increase in errors related to this system. Errors were primarily due to unscannable bar codes, mislabeled medications, the wrong medications being dispensed, and most commonly, nursing staff’s failure to scan. To assess the scope of the issue, we ran a point-of-care utilization report on users in one unit, uncovering remarkably low scan rates for both patient arm bands and medications. We had conducted an FMEA prior to initially employing BCMA; however, we never performed any postimplementation follow-up on the system. We decided to conduct another FMEA to pinpoint exactly what was causing the increase in errors and create workable solutions that addressed all facets of the process and all individuals involved.
Since we had multiple types of errors (eg, wrong drug, omission, wrong quantity, wrong patient, wrong label, extra dose) across disciplines and sites, we needed to conduct a comprehensive process review that could address all these errors and would bring together all the involved departments (ie, pharmacy, nursing, and IT). We chose an FMEA for two key reasons: while root cause analyses typically look at one type of error, FMEAs are designed to address multiple errors; and second, FMEAs can be performed on existing processes. Conducting an FMEA would allow us to identify and analyze points of failure and then design solutions to reduce risk.
Obtaining senior leadership’s support for this time-consuming process was a prerequisite to securing the needed time from staff. To gain leadership support, we informed them of the increase in errors related to BCMA. Routinely sharing our progress and successes throughout the lengthy FMEA process ensured their continued support.
Also important is establishing buy-in from end users, which is easier to accomplish when they are part of solution development. In turn, having end users on the team adds the perspective of those with everyday, hands-on experience. We asked nurse managers to identify the creative thinkers who could contribute practical solutions, and invited those nurses onto the team. To create an air of trust and openness, we made sure to dispel any fears of retribution for not previously following policy.
The Critical First Step: Process Mapping
The first and most important step in an FMEA is process mapping. This crucial step ensures all possible workarounds are addressed. Successful mapping requires attention to detail; it cannot be rushed. Also, the team must be clear that the task is to map what current users are actually doing, regardless of policy. Emphasizing that no disciplinary actions will be taken against those not following policy is necessary.
While we created separate process maps for pharmacy and nursing, all team members participated in mapping both disciplines. In pharmacy, mapping exposed these barriers:
In nursing, process mapping revealed these barriers:Handheld scanners varied in their bar code scanning abilities
Calculating Risk Priorities
Once process mapping was completed, we entered all identified failure modes into an FMEA documentation tool. The team then used a 10-point scale to score the severity, occurrence, and detectability of each failure mode, with 10 being the worst effect. The higher the detectability score, the lower the chance of catching the error before it reaches the patient. Multiplying the three scores gave us each failure mode’s risk priority number (RPN). While an RPN greater than 100 would normally be reviewed, the team decided on a cutoff of 250 because of the high number of failure modes (see Table 1).
Click here to view a larger version of this Table
Taking Action in Pharmacy
After all RPNs were assigned, the group discussed potential solutions. Solutions needed to be effective and sustainable, and the team focused on standardizing processes and maximizing technology. For pharmacy, action items included implementing standardized training for personnel working in the pre-pack room and standardizing pharmacist checking and quality assurance for pre-pack medications. We also decided that the door of the pre-pack room is to remain closed during packaging and checking to minimize interruptions and distractions.
The pharmacy automation team created a script for all pharmacy employees to use when contacted about a product not scanning (see Table 2). Pharmacy also developed and implemented a standardized process for identifying medications new to our system (ie, items for which the bar code is not recognized). The pharmacy buyer is responsible for identifying when a new product or new manufacturer arrives at the pharmacy. The product is then isolated until the new bar code can be added to the computer database. Pharmacy also collaborated with IT to add a second bar code label to IV syringes.
Click here to view a larger version of this Table
Since pharmacy processes were reviewed first, error-reduction solutions in pharmacy were enacted before the FMEA was completed. End users on the FMEA team communicated the effects of these positive changes, which helped pave the way for improvements in nursing.
Process Improvements in Nursing
Action items for nursing included developing and implementing a standardized process for troubleshooting BCMA (see Table 3). Additionally, we increased nursing accountability by reporting individual scan rates in the point-of-care utilization report for each nursing unit (although user names were removed). Seeing low scan rates made nurses more alert to following procedure. Nursing staff also was trained to use the task list on the electronic health record, which is updated immediately, rather than the printout, which may not have the most current data.
Click here to view a larger version of this Table
Nurse managers were trained to use the point-of-care utilization report and the bypassed five rights report with staff. From the point-of-care utilization report, nurse managers shared the individual scan rates with staff, conveyed the goal of 95% accuracy, and described the process improvements pharmacy made to help them meet that goal. Staff was given a grace period of 30 days to improve, after which remediation and disciplinary action could be pursued. During the grace period, each manager developed an action plan to improve scan rates and shared it with nursing leadership. The bypassed five rights report shows the length of time (in minutes) from when a medication was due to when it was actually scanned. This report also was shared with staff to encourage improvement.
Finally, IT was tasked with upgrading all handheld scanners to provide nursing staff with reliable equipment.
Three months after completing the FMEA, the team compared the before and after scan rates. We found significant improvements in the scanning of both the patients and the medications throughout the system (see Table 4). In addition, we have witnessed a culture change: nurses now become anxious if they cannot scan a product.
We learned valuable lessons from our FMEA process that should benefit any hospital embarking on a similar journey. First, FMEA can be conducted on a current process, but only if the climate allows for process mapping to match actual practices, not policy. Fear of retribution for not following policy can derail this crucial step. Investing the time that comprehensive process mapping requires is essential to the accuracy and completeness of results, and therefore to the success of the project. Second, building a team that includes end users to gain their buy-in is just as important as gaining the support of leadership. Third, implementing the most effective error-reduction techniques available is key. With more than 1500 nurses using BCMA at our hospital, standardizing processes and maximizing the use of technology in workflow were essential. Finally, FMEA is a lengthy, detail-oriented undertaking. Team members and leadership need to be kept abreast of progress and successes to minimize discouragement and maximize commitment and momentum.
While the commitment required to complete an FMEA is significant, the resulting process improvements and safety gains make this investment worthwhile. In fact, after sharing the project’s success with leadership, they have now tasked us with completing an additional FMEA for medication reconciliation.
Diane McClaskey, RPh, BCPS, is the pharmacy residency director and medication safety pharmacist for CoxHealth. A graduate of Drake University, she also serves as chair of the Missouri Society of Health System Pharmacists Web site committee. Diane’s current practice interests include critical care, medication safety, infectious disease, and cardiology.
Lisa Massey, RN, obtained her nursing degree from Burge School of Nursing in Springfield, Missouri, and has practiced nursing in both Texas and Missouri. She has been a registered nurse for 24 years and has a clinical background in critical care and legal nurse consulting. Currently, Lisa is employed in the office of patient safety as the medication safety facilitator for CoxHealth. Her current practice interests include medication safety, critical care, and medico-legal nursing.
Enter our Sweepstakes now for your chance to win the following prizes:
Just answer the following quick question for your chance to win:
Entries are limited to one entry per person in any active sweepstakes.
Thank you for your entry.